MFI Pump - Open Heart Surgery

[dicklague]

Yes,

we have idle and part throttle. I am going to convert those to AFR and try to do a test today. I am just finishing the MAP sensor installation just need to find 12vdc switched in the area to power the 5vdc power supply and I am done. Will keep you posted.

And......if any Pelicans have an iidea of the other readings in CO or AFR we are striving for or if there is a document....please speak up.

[jeffc280sl]

The Porsche spec doesn't require the use of a MAP sensor. You need to find a long stable incline where you can maintain 2400 rpms with the throttle on the 9 degree stop. When you have achieved this balance you have met the load spec so to speak. Your handbrake can help establish the load.

[dicklague]

I know it does not require MAP, but I want to see if I can correlate required load to map then I can know where I am and monitor it on an almost continuous basis and see what happens under varying conditions.

[jeffc280sl]

Sounds like a fine idea. Others may like to know the load expressed in terms of inHg.

[dicklague]

I have been fiddling with new Daytona Sensors Wego IV AFR gauge/data recorder with some pretty good results.

I have converted the CO readings to AFR and have come up with the following:

1. IDLE 900 RPM 13.56 AFR +/- .5% which would be +/- .06 AFR

2. 9 Degrees of throttle opening @2400 RPM under load in second gear should be
13.56-13.38 AFR for the part throttle setting.

I started out with going way to lean on both setting and it was very very lean. I ended up going 3 clicks richer on part throttle and a 2 clicks richer on idle and it made a big difference. Just a little too rich. I am going to try on click leaner on each of them.

I am still working on correlating map. I have to calibrate the chart to actual voltage at rest and work that in. I will post charts as soon as I have this all done.

So far I am very please with how close this AFR gauge has allowed me to get. Very impressive.

[jeffc280sl]

It's great to be able to take data versus tuning by ear or seat feel under acceleration.

If you can't find a conversion formula in your data logger you can make a custom table with a mity vac brake bleeder and vacuum gauge. Create various vacuum levels and read the voltage.

[dicklague]

That is a great idea. I may just do that if I can find my vacuum pump. What I did is take the voltage at ambient air pressure which was 4.7 VDC and used that for the max. I am also going to look up the specs for the GM 1bar sensor I am using. I think that will give max and min. I know I have 4.99 volts regulated going into the MAP sensor.

Dick

[jeffc280sl]

All you need is a kpa/inHg calculator. Check it with your mity vac to make sure.

Home » Shop » Sensors » Pressure Sensors » Genuine Delphi / GM 1 Bar Map Sensor - 16254719 - 16175339 - 12247561 (96 case buy only)

[dicklague]

I am getting some good results with the Wego IV AFR instrument. It has really helped me tweak the MFI using A/F ratio and know exactly where I am going. The data logging is terrific. I logs up to 1.5 hours and that is plenty to get the engine fully warm and see exactly where you are. It is easy to download and study the chart. I have also been importing the raw text file in to a spread sheet program so you can study every half second. This can give you up to 8000 data readings so there is a lot of data. I am finding it handy to go line by line and compare RPM, AFR, and MAP. It gives you a precise reading like a dyno run, but you can do it on the street or on the track.

Here is a .JPG of a 5 minute slice of a chart from a run to day on Stunt Rd. and Mulholland.




This shows rpm in RED and AFR in Green. time is along bottom. AFR right and RPM on left scale. this is only 5 minutes. You can scroll through the whole run. The instrument is always recording. It adds new data to the end. You can erase everything in the memory for a new run, or if you forget it just records the latest data and adds to end of the file.

I am very happy with the way the MFI in my 1973 is running now.

[dicklague]

Here is a spread sheet of small portion of the run today. this was wide open throttle for a short straight on Mulholland. You can see 6848 RPM with 13.88 AFR and 4.35 inches of mercury.



Give me a good idea that WOT is OK and not too lean.......it sure felt and sounded good!!

[dicklague]

the last column is actually MAP voltage. 30 inches of mercury is 4.7 volts in this set up. I will work on a formula to convert this spread sheet to IN hg.

[Mr Beau]

Best power is around 12.5 to 13.2 so you may want to richen things up a bit more (assuming the readings are accurate). You'll also lower EGT's as you move away from 14.7.

It's great seeing modern technology make MFI work again as Porsche originally intended!

[dicklague]

M. Beau,

After looking at those numbers again I do agree with you. I need to try another click richer on part throttle/mid. It is a lot better than it was......started way too lean.

I did a free air calibration the day before, so it is +- .5% accurate. The weather was cool and it possible that more running would have raised temps in the engine.

I converted some of the Porsche MFI data from CO to AFR and they are calling for 3.2 to 3.6.....but I agree with you and would rather be a bit rich than too lean. Richness helps cooling as well in a air cooled engine.

here is the unit I am using:



It is mounted under the dash where the ashtray used to be mounted. The mini USB port is on the front so you can download to a laptop.

[jeffc280sl]

Looking good. I would try some smoothing on the graphs. Stretch out the time during WOT. I like to look at throttle response. Drive in 3rd or 4th gear at 3k rpms nice and smooth to get a good starting point. Then mash the accelerator. What happens to afr when the MAP drops? Obviously there is a big gulp of air being taken which can make the afr go momentarily lean. You want afr to respond as soon as possible and during acceleration be almost flat between 13 and 13.7.

[jeffc280sl]

I think it's interesting to note that afr bounces all over the place while driving. Reading spark plugs is better than nothing but fine tuning the fip to meet a performance/emissions spec can only be done on a dyno or with a portable afr meter.

[dicklague]

Here is a 1 minute picture with that RPM peak in the center. This one is AFR and RPM




And the one below is the INhg and AFR chart for 1 minute.

[jeffc280sl]

Can you just capture the perios between 3540 and 3560 seconds? Can you add map to afr and rpm?

I believe your map voltages are upside down. You should have very low pressure during acceleration.

[Mr Beau]

I also think the MAP looks upside down. If you can't plot more than 2 sets of data on the same graph, why not dump it all into Excel (or OpenOffice) and look at the data that way? It's tough to tell for sure but it doesn't seem like the MAP and RPM traces line up the way that they should.

[dicklague]

Yes, you can easily dump it into Excel and look at it that way. By the way the MAP is expressed in inches of mercury. So 30 is atmospheric and lower readings are more vacuum. This is NOT voltage on this chart. When you do Excel you get RPM, AFR, and MAP voltage.

Is this confusing?

[jeffc280sl]

I understand that less than atmospheric is vacuum and that it is expressed as inHg. The MAP sensor outputs 0 to 5V to the af meter. What I'm saying is your voltage table is upside down. Your chart shows map at 3535 sec moving towards atmospheric pressure (rising) when in fact map pressure is lower (falling) under acceleration.

In your mind zero the atmospheric pressure. Anything less than 0 (30 inHg) is negative air pressure or vacuum. As vacuum pressure increases map pressure reduces. If you increase vacuum pressure by 5 inHg (think of your mity vac) the relative map pressure is 30 minus 5 or 25 inHG.